Abstract
Purpose :
To evaluate retinal vessel density within various retinal structural layers using optical coherence tomography angiography (OCTA).
Methods :
In this IRB-approved study, 22 normal eyes (from 22 subjects) were imaged using the Spectralis OCT2 with prototype OCTA acquisition software (version 6.9.2.700). Macula OCTA scans were captured at isotropic resolution of 768 A-Scans per 15 degrees. Three scans were obtained, one foveal-centered 15x15 degrees scan and two 15x5 degrees scans displaced temporally and nasally by 15 degrees along the Fovea – Bruch’s membrane opening (BMO) axis. Projection artifact removal (PAR) software was used to allow quantitative assessment of deeper retinal layers. Vessel density (i.e. amount of flow signal) within each retinal nuclear and plexiform layer was assessed across the scan and its average over all subject eyes was plotted as a vessel density profile over this Fovea-BMO axis (with the fovea assigned a position of zero on the axis, nasal regions as negative distance, and temporal regions as positive). Vessel density was correlated against the retinal layer thickness at the corresponding locations using Spearman correlation.
Results :
Of the 22 normal subjects, 14 were male and 8 were female with a mean age of 34 ± 7. For the nerve fiber layer (NFL), the vessel density was highest nasally and declined towards the fovea and was near zero temporal to the fovea with or without PAR (Figure). For all other retinal layers, the retinal vessel densities were greatest in the parafoveal retina, peaking approximately 5 degrees from the foveal center. Before PAR, the temporal parafoveal vessel density was highest in the inner plexiform layer (IPL). Following PAR, the vessel density in the IPL showed the most dramatic decrease. Following PAR, the ganglion cell layer (GCL) demonstrated the highest temporal parafoveal density with an increase in the peak temporal parafoveal GCL/IPL vessel density ratio from 0.78 before PAR, to 1.40 after PAR. The vessel density was correlated with retinal layer thickness (r = 0.432 to 0.511; P<0.05 among the various layers).
Conclusions :
Retinal vessel density varies significantly among the various layers, with significant regional variability. Projection artifact can significantly impact retinal vessel density in the deeper layers, but the effect appears to be most pronounced in the IPL.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.